Based On The Nonlinear Theory Of Chinese Phonetic Analysis And Prediction

At present, the analysis and prediction of speech signal are all using linear theory and linear prediction technique, but the speech production system is complicated nonlinear and has chaotic property as well as fractal feature, so linear methods are inadequate. Therefore, the nonlinear characteristic of Chinese speech are further studied, combined with Radical Basis Function Network(RBF Network for short), a nonlinear predictor is designed.Firstly, theoretical basis of speech signal nonlinear prediction and prediction tools are analyzed, and methods of solving phase space reconstruction parameters containing delay time、embed dimension are further studied, which are firstly solved by C-C algorithm, according to the limitation of results, then combined with auto-correlation algorithm and FNN(False Neatest Neighbors) algorithm are solving respectively. According to select sample rate at experimentations, statistical method is used to study. The results show that sample rate has little influence on delay time and embed dimension.Secondly, based on nonlinear theory, nonlinear characteristic parameters of Chinese speech phonemes are studied. The maximum Lyapunov components are solved by Wolf-algorithm and correlation dimension are solved by GP-algorithm, which indicate Chinese speech has chaotic characteristics.At last, based on nonlinear characteristics of Chinese speech signal, Radical Basis Function (RBF) network analysis methods are applied to design nonlinear predictor. The averages of the delay time for Chinese speech phonemes determine the neurons number of the input layer and output layer for RBF neural network model, and The averages of embedding dimension determine the neurons number of the hidden layer, The simulation results indicate:Compared with the linear predictor, prediction error of nonlinear predictor based on RBF network is significantly decreased and has higher performance as well as prediction accuracy.